Please refer to the following article for tips on how to improve model performance.
Key ideas:
- Use the Performance Advisor to automatically identify areas of improvement.
- Consider the workflow context in which you are using your model and consider workflow improvements such as accelerator, model reference, Fast Restart, parallel builds, etc.
- Use Solver Profiler (for variable step solver only) to locate blocks that are triggering solver exceptions, resets, and zero crossings. Consider replacing these blocks with alternatives. Consider the following common example: Why you should never break a continuous algebraic loop with a Memory block » Guy on Simulink - MATLAB & Simulink (mathworks.com)
- Use the Simulink Profiler to identify blocks that are taking the majority of the simulation time. Consider replacing them with blocks of lower fidelity or running them at a reduced sample rate. It's not always a trade-off, similar to other programming languages, sometimes a change in algorithm can achieve the same fidelity with improved speed.
